food and bioproducts processing 9 1 ( 2 0 1 3 ) 396402

Contents lists available at ScienceDirect

Food and Bioproducts Processing

journal homepage: www.elsevier.com/locate/fbp

Effect of separation and thawing mode on block

freeze-concentration of coffee brews
F.L. Moreno a , C.M. Robles a , Z. Sarmiento a , Y. Ruiz a , J.M. Pardo b,
a
b

Agroindustrial Process Engineering, Universidad de La Sabana, Km 7 Autopista Norte de Bogot, Cha, Colombia
School of Engineering, Universidad Panamericana, Campus Mxico, Augusto Rodin 498, 03920, Mxico, D.F., Mexico

a b s t r a c t
Coffee brews were freeze-concentrated using block technique to evaluate the effect of thawing temperature, thawing
mode and separation mode on solute yield. Samples were frozen and solute recovered using different thawing and
separation conditions. Three thawing temperatures (20, 4 and 1 C), two thawing modes (microwave assisted and free
thawing) and two liquid fraction separation modes (gravitational and vacuum assisted) were tested. Solute yield was
evaluated as a function of thawing ratio. Data were tted to a monomolecular model to compare separation quality
by means of the solute recovery ratio for each treatment. A combination of microwave assisted thawing and vacuum
separation showed the best results in terms of solute recovery ratio. On the other hand, applying microwave assisted
thawing or vacuum separation individually did not signicantly enhance the solute recovery ratio. Additionally
a thawing temperature effect on the solute recover ratio was also found. These results show that it is possible to
improve the separation quality of Block FC by combining this technology with vacuum-assisted separation microwave
heating. These results suggest that block freeze-concentration has potential industrial application.
2013 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
Keywords: Cryoconcentration; Thawing; Coffee; Vacuum; Microwave; Solute yield; Monomolecular model

1.

Introduction

Freeze-concentration (FC) is a technology used to remove

water at low temperatures preserving the quality of the
original material (Miyawaki et al., 2005). Water remotion is
achieved once water ice crystals are formed, leaving behind
a more concentrated uid (Snchez et al., 2009). Moreover, FC
allows to reduce processing time during the industrial freeze
drying (Boss et al., 2004).
FC techniques can be classied in three groups: suspension
crystallization (Huige and Thijssen, 1972), lm freeze concentration (Snchez et al., 2009) and freeze-thaw method, known
also as block freeze concentration (Aider and de Halleux,
2009; Snchez et al., 2011a). For suspension crystallization a
scraped surface heat exchanger is used to form a suspension
of ice crystals that can be subsequently separated (Miyawaki
et al., 2005; Habib and Farid, 2006). In Film FC a single crystal layer is formed by contact with a refrigerated surface
(Ravents et al., 2007; Snchez et al., 2011b). Freeze-thaw FC

can be achieved using two different techniques: total or partial

block (Nakagawa et al., 2010a). In the partial block technique,
the solution is partially frozen and the remaining concentrated liquid is separated after a pre-determined time (Burdo
et al., 2008). On the other hand, the total block technique
consists of completely freezing the sample with a subsequent thawing until the desired amount of solute has been
recovered (Aider and de Halleux, 2009). Consequently, block
FC comprises three important steps: freezing, thawing and
separation.
Increasing ice purity is a big challenge for block FC (Aider
and de Halleux, 2008). Although, it is clear that the mass
of the collected liquid increases with the collection time,
increasing it until collection of the entire original solution
will lead to no overall concentration effect. Therefore, it is
important to determine the mass of the collected sample in
which solute recovery and concentration are high enough
in order to stop the process at this point. In the same
way, it is important to establish the thawing conditions or
aids in separation stage to recover as much solute as possible.

Corresponding author. Tel.: +52 54821600.

E-mail addresses: jpardo@up.edu.mx, jmauriciopardo@gmail.com (J.M. Pardo).
Received 14 June 2012; Received in revised form 14 January 2013; Accepted 18 February 2013
0960-3085/$ see front matter 2013 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.fbp.2013.02.007

397

food and bioproducts processing 9 1 ( 2 0 1 3 ) 396402

Fig. 1 Experimental setup.

The block FC method has been recently studied as a way
to improve solute yield (Gao et al., 2009; Okawa et al., 2009;
Nakagawa et al., 2010a; Miyawaki et al., 2012). Furthermore,
microwave assisted and gravitational thawing have been studied for milk whey (Aider et al., 2008) and for maple sap (Aider
and de Halleux, 2008). It is worth noting that there are no
reported studies about vacuum application and its relationship with process variables during the thawing and separation
stages of block FC.
Meanwhile, coffee is the most traded food commodity
worldwide (Esquivel and Jimnez, 2012). The functional and
organoleptic properties of coffee are highly important for its
processing and trading. Coffee processing technologies play a
large role in dening the quality of the beverage (Jot et al.,
2010), therefore, in the soluble coffee industry, technologies
that preserve the initial characteristics of the brew are useful.
Currently, suspension FC is the unique technique that is being
used by coffee industry. This technique shows a high efciency
and purity of separated ice crystals, however, it is recognized
as an expensive method for food concentration (Aider and de
Halleux, 2009). Therefore, the use of this separation process
is restricted to large-scale processing plants. For this reason,
the block FC technique has been proposed in order to reduce
cost and maintain high crystal purity (Aider and de Halleux,
2009), including different modes to increase solute recovery in
thawing and separation stages, looking forward to implement
this technique at different industrial scales.
In this paper, the total block FC of coffee brews is studied,
analysing the effect of separation mode (gravitational and vacuum assisted), thawing temperature (1 C, 4 C and 20 C) and
thawing mode (natural convection and microwave assisted)
on solute yield.

the precision of refractometer (0.5 Brix), uncertainty in solid

fraction measurement corresponded to 0.00438. It is clear
from other reports that the initial concentration affects the
results of freeze concentration (Aider et al., 2008). However, in
these experiments initial solid content was used as a constant
parameter for the experiment to observe the effect of the other
factors.

2.2.

Methods

2.2.1.

Freeze concentration protocol

Samples were frozen in cylindrical containers (200 mL) at

12 C and then stored during 72 h at the same temperature. Afterwards, samples were thawed and the concentrated
fraction was separated under different conditions. As shown
in Fig. 1, frozen samples (1) were placed in a chamber with
controlled temperature (5) and drops of liquid fraction were
collected by gravity in an external collection vessel (2). The
chamber consisted of an upright freezer with thermostat
(Haceb 87L, Colombia) for the two lowest temperatures and an
incubator chamber (Selecta Celmag-L, Spain) for the highest
temperature level. Below the vessel was a scale (3). (Mettler Toledo AB 204 S, Switzerland. Precision 0.1 g), used to
record weight changes during thawing. The collection vessel
was removed approximately every time that 7% of liquid was
collected in order to measure the liquid concentration. The
thawing procedure was performed at three different thawing
temperatures (TT) xed in the chamber (1 C, 4 C and 20 C
1 C), two thawing modes (TM) and two separation modes
(SM) as shown in Table 1 following a complete factorial design.

Table 1 Experimental design.

2.

Materials and methods

2.1.

Materials

Coffee solutions were prepared from commercial soluble coffee, (Aroma brand, Colombia) and distilled water at 20 C. An
initial concentration of 4.3 g of dissolved solids/100 g of solution (5 Brix) was used. Concentration was measured using
a refractometer (Abbe refractometer model RL3 Polskie Zaklady Optyczne Warszawa, Poland), together with a calibration
equation in order to account for the deviation caused by the
mixture of solids present in coffee. The calibration equation
was obtained by preparing solutions at 10, 20, 30, 40 and
50 Brix and measuring their solid contents using an oven drying procedure at 103 C for four hours according to (NTC 4602,
1999). Measurements were performed in triplicate. The following equation was obtained: Xs = 0.0087 Brix (R2 = 0.99). Due to

Treatment

1
2
3
4
5
6
7
8
9
10
11
12

Thawing
temperature
( C) (TT)

Thawing
mode (TM)

20
4
1
20
4
1
20
4
1
20
4
1

+1: microwave assisted thawing.

1: free thawing (without microwaves).

1
1
1
+1
+1
+1
1
1
1
+1
+1
+1

Separation
mode (SM)
Gravitational
Gravitational
Gravitational
Gravitational
Gravitational
Gravitational
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum
Vacuum

398

food and bioproducts processing 9 1 ( 2 0 1 3 ) 396402

Thawing and separation modes are explained in Sections 2.2.2

and 2.2.3. All experiments were performed in triplicate.

Additionally, Y can be obtained as shown in Eq. (4) by combining Eqs. (1)(3):

2.2.2.

Y = CF f

Thawing mode

For some samples, as it is shown in Table 1, the thawing stage

was initiated by applying a microwave pulse (+1) and for others
thawing was freely developed under a settled external temperature (1). Therefore, half of samples were exposed to 5.1 kJ
of microwave energy that was delivered in one minute by a
microwave oven (Haceb Arezzo-0.7, China). This energy pulse
was made of three short pulses of 2 s at 1.7 kJ, leading to an
average power delivery of 85 Watts during one minute. Afterwards, the thawing and separation procedure was carried out
at the settled thawing temperature following the procedure
described in Section 2.2.1.

2.2.3.

2.3.

Data analysis and modelling

2.3.1.

Thawing fraction

A thawing fraction (f) was dened to follow the development of

the process. f was measured as the ratio between thawed mass
and that of the original solution, dened by Eq. (1) (Nakagawa
et al., 2010b; Miyawaki et al., 2012):
(1)

m0

where, f is the thawing fraction, mliq is the collected liquid

mass, m0 is the initial mass.

Solute yield

Solute yield is dened as the fraction of recovered solute, calculated by the relation between mass of solids present in the
separated liquid and the mass of solute present initially in the
original solution (Nakagawa et al., 2010a):
Y=

ms liq
ms 0

(2)

where, Y is the solute yield, ms liq is the solute mass in liquid

fraction, ms 0 is the initial solute mass.

2.3.3.

Concentration factor

A concentration factor was dened as the relation between

the solid concentration in the liquid fraction and the solid
concentration in the initial solution:
CF =

If Y = f

ms liq
ms0
=
m0
mliq

(5)

It can be inferred from the Yf plot that the further a point

is from the diagonal line, the higher the concentration of
the sample. It is worth pointing out that the behaviour of
the FC process in this plot is a curve bounded between the
points (0,0) and (1,1). This relationship can be represented by
a monomolecular model Eq. (6), which is frequently used in
the statistical analysis of nonlinear relationships (Nothnagl
et al., 2004).
dY
= r (1 Y)
df

Xs liq
Xs 0

(3)

where, CF is the concentration factor, Xs liq is the solid fraction

in freeze concentrated liquid fraction, Xs 0 is the solid fraction
in initial solution.

(6)

where, Y is the solute yield, f is the thawing fraction, r is the

solute recovery ratio (constant for each condition).
Eq. (7) is obtained by solving the differential equation and
transforming it into a linear equation using natural logarithms:
ln

mliq

2.3.2.

Mathematical model

Both solute yield (Y) and thawing fraction (f) vary between 0
and 1. It is important to note that a diagonal line with slope
equal to one in the plot Y against f represents a process without
concentration and is considered a process with a perfect solute
inclusion (Nakagawa et al., 2010a). Therefore, at any point on
the diagonal line, the concentration of solids in the portion
sample is that of the initial sample as seen from Eq. (5).

Separation mode

Two different separation modes were carried out: gravitational

and vacuum assisted. The rst consisted of drop collecting
under gravity without any interference. In the second, a
vacuum pressure of 1 kPa (74.7 kPa abs) was applied using
an aspirator (Bchi model B-169, Switzerland) to route and
increase the movement of the liquid fraction that was trapped
inside the frozen portion. Thawing and separation procedures
were carried out as described in Section 2.2.1. The vacuum
pump was disconnected in order to measure mass and solid
concentrations and reconnected after collecting the sample.

f =

2.3.4.

(4)

 1 
1Y

=rf

(7)

Therefore, the constant r is the slope of straight line calculated from Eq. (7) and is not directly related with time.
Furthermore, it represents the solute yield per thawed liquid
fraction and can be named solute recovery ratio. The higher
the slope r, the less ice that should be melted to recover a
dened amount of solute. Thus r can be used as an indicator
of separation quality. The monomolecular model is an intrinsically linear model with one parameter (r). As CF is the ratio
between Y and f according to Eq. (4), the r parameter can be
understood as the change of concentration factor in the graph
Y vs. f.

2.3.5.

Statistic analysis

Parameters of the monomolecular model for each treatment

were obtained by a simple linear regression procedure using
SAS 9.2. The result was the r value for each of twelve evaluated
treatments. A ratio comparison with condence interval procedure at p < 0.05 was used to determine differences of r value
between treatments.
Additionally, it is possible to identify the inuence of factors TT, TM and SM on the solute recovery ratio (r). The r
value can be considered as a functional value, it represents
the relationship between Y and f. A multiple linear regression
procedure (p < 0.05) was applied using SAS 9.2 to this purpose.
The signicance of the mean and combined effects on the r
values was established.

399

food and bioproducts processing 9 1 ( 2 0 1 3 ) 396402

Table 2 Solute recovery ratio.

TT ( C)

Treatment
number
11
10
12
6
9
3
1
4
7
2
5
8

4
20
1
1
1
1
20
20
20
4
4
4

TM
MW
MW
MW
MW

MW

MW

SM

Solute recovery
ratio(r)

Vacuum
Vacuum
Vacuum
Gravitational
Vacuum
Gravitational
Gravitational
Gravitational
Vacuum
Gravitational
Gravitational
Vacuum

6.74
6.42
6.26
5.61
5.60
5.12
5.01
5.00
4.74
4.32
3.37
3.04

R2

RMSE

0.978
0.976
0.931
0.979
0.980
0.943
0.989
0.977
0.990
0.984
0.968
0.963

0.078 a
0.117 a
0.216 a
0.052 b
0.047 c
0.419 d
0.031 e
0.117 e
0.018 e
0.041 f
0.106 g
0.030 h

Treatments with the same letter do not differ signicantly.

3.

Results and discussion

The solute recovery ratio (r), dened in Eq. (7), was obtained for
each treatment and these values are presented in decreasing
order in Table 2. In this table, the best conditions for solute
recovery in freeze concentration correspond to those with
the highest r values. There, it can be seen that regression
coefcients vary between 0.93 and 0.99, showing a good t to
the model. The majority of treatments were statistically different from each other, except for the treatments 10, 11 and
12 that showed statistically the same values (95% condence)
as well as treatments 1, 4 and 7.
Fig. 2 shows the comparison between the Y values that
were estimated using Eq. (7) and Y values obtained from
experimental data. It is clear that the data points were distributed close and around the continuous line that represents
(Y = Y). Analyzing the slope of predicted and experimental data
line (m = 0.999) and regression coefcient (R2 = 0.94), it can be
inferred that predicted values tend to be similar to experimental values. This result conrms the good t of experimental
data to the monomolecular model.
At this point, it is important to highlight the usefulness
of parameter r in FC quality analysis. As discussed earlier,

1.0

predicted values

0.8
0.6
0.4
0.2
0.0
0

0.2

0.4
0.6
0.8
Y experimental values

Fig. 2 Parity plot: experimental Y values against predicted

Y values.
Regression line obtained: y = 0.999x; R2 = 0.94.

the r value represents the amount of sample that needs to be

melted in order to recover a dened amount of solids. It can
be seen from experimental data (Table 2), that solute recovery
ratios vary from 3.04 to 6.74. Thus, these r values are showing
that under some experimental conditions (treatment 8), it is
necessary to thaw nearly half of the sample in order to recover
75% (Y = 0.75) of the solids. On the other hand, in the best case
scenario (treatment 11), it will be necessary to thaw only 20%
of the sample in order to recover the same amount of solute.
Furthermore, the concentration of recovered liquid Xliq can be
calculated using r values and, Eqs. (1), (2) and (7).
Xliq =

Y ms0
Y X0
=
f m0
f

(8)

where, X0 : initial solids concentration in the sample.

If Eq. (7) is included the resulting expression is a function
of r:
Xliq =

Y X0 r
ln (1/(1 Y))

(9)

Therefore, using experimental data which have r values

between 3.04 and 6.74, it can be estimated that to recover 75%
of solids, the concentration of solids in the liquid samples
will vary between 0.08 and 0.18 depending on the procedure selected. This information, combined with information
on mass transfer speed, is a useful tool to dene block FC
processing strategies. Therefore, modelling heat and mass
transfer kinetics should be part of the future work.
A regression analysis was used to quantify the effect of
each parameter on the solute recovery ratio (Table 2). The
experimental design selected for this work, made it possible to obtain linear main effect and linear combined effect
of the evaluated parameters. These results are presented in
Table 3. Two parameters showed a signicant effect on solute
recovery ratio: thawing temperature (both TT and TT TT)
and the combination of separation mode with thawing mode
(SM TM).
The thawing temperature had a signicant effect on solute
yield as illustrated in Fig. 3. The trend was not linear because
of the signicant inuence of TT TT (Table 3). Therefore,
TT = 1 C is better than TT = 20 C and better than TT = 4 C for
all separation and thawing conditions where the combination
of MW-Vacuum is not present.
The signicance of thawing temperature can be explained
by changes in ice structure. The solute recovery depends on

400

food and bioproducts processing 9 1 ( 2 0 1 3 ) 396402

Table 3 Effect of TT, TM and SM on solute recovery

ratio.
Pr > |t|

Standard estimator

Intercept
TT
TM
SM
TT TT
TT TM
TM TM
SM TT
SM TM
SM SM

6.028
0.608
0.310
0.186
0.027
0.018

0.003
2.184

<0.0001
0.0028
0.6110
0.7595
0.0023
0.6586

0.9348
0.004

the capacity of the separation system to collect concentrated

liquid fractions and this is related to the porous structure
and density of frozen matrix. Water solutions experience
density changes during freezing and thawing (Akyurt et al.,
2002). Freezing and thawing temperatures inuence freezing
and thawing speeds, affecting the ice porous structure (Pardo
et al., 2002) and the movement of liquid through it. Moreover,
if diffusion of liquid is slower than heat transfer, some ice
would be expected to melt and to dilute the solution before it
can be recovered. Therefore, the rate of mass transfer affects
the solute recovery ratio because at low mass transfer rates,
greater portions of the sample should be melted in order to
obtain a certain amount of solute. Thus, for this kind of freeze
concentration process it is important to nd a good balance
between thawing speed (heat transfer) and separation speed
(mass transfer) not only to optimize processing rate, but also to
nd the best recovery condition. Meanwhile, TT did not show a
linear effect, which was not expected, therefore, further work
in order to relate changes of frozen sample s microstructure
with processing parameters can lead to a deeper understanding of block FC.
It was seen from experimental data (Table 2) that, the highest solute recovery ratio was obtained for treatments 10, 11

0.8

0.6
1 model
1 exp
4 model
4 exp
7 Model
7 Exp
10 model
10 exp

Parameter

0.4

0.2

0
0

0.2

0.4

0.6

0.8

f
Fig. 4 Effect of separation mode on solute yield
treatments numbered according to Table 1 (1,
4 = gravitational; 7, 10 = vacuum) Lines are modelled by Eq.
(7) and parameters in Table 3. Each symbol corresponds to
data collected from triplicate experiments.
and 12, in which a combination of microwave thawing and
vacuum separation was used. Moreover, temperature did not
show a signicant effect. This observation is in accordance
with the results of regression analysis (Table 3) in which the
combined parameter had a bigger standard estimator than TT
and TT TT parameters. Other parameters such as TM and SM
had no signicant inuence when analyzed alone. A similar
result, where MW thawing did not show a signicant effect on
separation efciency was reported by (Aider and Ounis, 2012).

1
1

0.8
0.8

0.6
3 model
3 exp
6 Model
6 Exp
9 model
9 exp
12 model
12 exp

0.6

0.4

1 model

0.4

1 exp
2 model

0.2

2 exp

0.2

3 model
3 exp

0
0

0.2

0.4

0.6

0.8

f
Fig. 3 Effect of temperature on solute yield treatments
numbered according to Table 1 (1, 20 C; 2, 4 C; 3, 1 C).
Lines are modelled by Eq. (7) and parameters in Table 3.
Each symbol corresponds to data collected from triplicate
experiments.

0.2

0.4

0.6

0.8

f
Fig. 5 Effect of thawing mode on solute yield treatments
numbered according to Table 1 (3, 9 = free; 6,
12 = Microwave). Lines are modelled by Eq. (7) and
parameters in Table 3. Each symbol corresponds to data
collected from triplicate experiments.

401

food and bioproducts processing 9 1 ( 2 0 1 3 ) 396402

Table 4 Thawing time and concentration factor.a

Treatment
10
4
7
1
11
5
8
3
2
9
12
6
a

TT ( C)

TM

20
20
20
20
4
4
4
1
4
1
1
1

SM

1
1
1
1
1
1
1
1
1
1
1
1

Thawing time (h) at f = 0.5

Vacuum
Gravitational
Vacuum
Gravitational
Vacuum
Gravitational
Vacuum
Gravitational
Gravitational
Vacuum
Vacuum
Gravitational

2.7
3
3
3.5
12.5
25.6
31
35
48.3
50
75
80

Y at f = 0.5
0.98
0.96
0.90
0.94
0.94
0.84
0.72
0.93
0.94
0.93
0.95
0.94

0.03
0.05
0.12
0.07
0.06
0.07
0.08
0.05
0.07
0.04
0.13
0.04

CF at f = 0.5
1.96 0.05
1.92 0.05
1.8 0.11
1.88 0.05
1.88 0.15
1.68 0.11
1.44 0.17
1.86 0.2
1.88 0.05
1.86 0.55
1.9 0.34
1.88 0.55

Values in increasing order of thawing time. Average values with standard deviation.

Figs. 46 depict this situation, in which the effect of microwave

thawing and vacuum separation is signicant only when both
of them are present in the treatment.
Experimental results showed that the r value was increased
up to two times when MW was combined with vacuum separation. This means that a similar Y can be reached with a
lower f, increasing concentration of separated liquid portion.
On the other hand, the effect of vacuum was very small without microwave assisted thawing as seen in Fig. 5. Actually,
r values decreased when vacuum was present, but Table 3
shows that this effect was not signicant when compared
with other parameters. This suggests that vacuum separation promotes the separation of concentrated liquid fraction
from the ice matrix and increases process efciency only when
microwave thawing is present. Therefore, this synergy when
MW and vacuum are combined shows that this operating condition is an appropriate tool to improve separation quality in
block FC.

0.8

0.6
2 model
2 exp
5 Model
5 Exp
8 model
8 exp
11 model
11 exp

0.4

0.2

0
0

0.2

0.4

0.6

0.8

f
Fig. 6 Effect of thawing mode on solute yield treatments
numbered according to Table 1 (2, 8 = free; 5,
11 = Microwave). Lines are modelled by Eq. (7) and
parameters in Table 3. Each symbol corresponds to data
collected from triplicate experiments.

This synergistic behaviour can be explained using two

different mechanisms: accelerated internal thawing and
hydrodynamic movement of molten portions of the sample.
As it is known, the absorbed microwave energy is related
to the loss factor of the material. Thus, substances with a
high loss factor will readily absorb microwave energy, while
those with a low loss factor are either reecting or transparent to microwave energy (Wu et al., 2004; Wang et al.,
2011). It is important to highlight that the loss factor of liquid
water is greater than that of ice (Ryyniinen, 1995). Microwaves
can penetrate the frozen layer and reach the unfrozen liquid
(Rattanadecho, 2004). Therefore, it would be expected that during a microwave pulse, liquid pockets will be formed in regions
with higher concentration of solids and lower melting points.
However, these liquid pockets cannot be easily separated from
the sample because of the tortuosity of the paths inside the
frozen matrix. Here, vacuum plays a synergistic role because
pressure difference helps evacuating the concentrated liquid phase easily promoting liquid movement and separation.
Experimental data showed that this hydrodynamic movement
enabled by vacuum is a good way to accelerate liquid removal.
Therefore, in a future paper, the effect of these parameters on
the Block FC process kinetics will be analyzed.
As solute yield indicates the amount of solute recovery but
not the process velocity, it is useful to combine this parameter with the analysis of thawing time. The thawing time (t)
required to collect 50% of liquid (t at f = 0.5) is shown in Table 4
in increasing order. Also, the solute yield (Y) and the concentration factor (CF) at f = 0.5 are presented.
The thawing time was lower for treatments 10, 4, 7, and 1,
which correspond to TT, 20 C. Treatments with TT, 4 C and TT,
1 C resulted in a 25-fold increase in thawing times. The result
is similar for the solute recovery rate, in which TT showed a
signicant effect. Comparing treatments with the same TT,
microwave thawing slightly reduced thawing time.
Vacuum separation helped to reduce t slightly. At the same
time, vacuum separation allowed more solute to be recovered.
This suggests that vacuum separation promotes the separation of the concentrated liquid fraction from the ice matrix and
increases process efciency, showing this separation mode as
an interesting topic in block FC.
The concentration factor (CF) at 50% of thawing is presented for each condition in Table 4. Values between 1.4 and
1.9 were obtained. These values are comparable with those
obtained using other concentration techniques. For example, in falling lm FC, CF is around 2 for one stage of FC
(Ravents et al., 2007), in progressive FC, where CF is around

402

food and bioproducts processing 9 1 ( 2 0 1 3 ) 396402

2.8 (Miyawaki et al., 2005), in Block FC, CF = 1.4 without separation assistance (Nakagawa et al., 2010b) and in microwave
assisted thawing Block FC, CF is close to 1.7 (Aider and Ounis,
2012) for one stage.
The highest CF was obtained for treatments with vacuum
separation, due to the positive effect of pressure difference on
the movement of the concentrated liquid fraction.
On the other hand the shortest thawing time and highest solute recovery rate was obtained for treatment 10 (TT,
20 C, MW thawing and vacuum separation). This condition
gave the best overall is appropriate for the performance of FC,
recovering 98% of solute when 50% of mass is thawed and just
spending 2.7 h of thawing time.

4.

Conclusions

The relationship between solute yield and thawing fraction

can be represented using the monomolecular model. Linearization of this model, using slope r which is not related to time,
simplies the analysis in terms of the separation efciency,
and quality of the recovered liquid.
The combination of microwave assisted thawing and vacuum separation exerts a synergistic positive effect on solute
recovery ratio. Experimental results showed that this operating condition is an appropriate tool to improve separation
quality in block FC.
Thawing temperature has a signicant effect on solute
recovery ratio; both TT and TT TT, are highly inuential
parameters. Therefore, under the experimental conditions
used in this work temperature did not show a linear effect
on separation quality.
Block freeze-concentration is a potential technology for
food concentration. Process alternatives such as, thawing temperature control, microwave thawing and vacuum separation
are helpful to increase solute recovery and should be taken
into account when scaling up this technology.

Acknowledgements
Authors thank to Dr. Edgar Benitez for his statistics consulting.
Author Moreno F. L. thanks to COLCIENCIAS for supporting
his doctoral research by project 1230521-28461. Author Ruiz Y.
thanks to COLCIENCIAS for its condonable grant for doctoral
studies (2004).

References
Aider, M., de Halleux, D., 2008. Passive and microwave-assisted
thawing in maple sap cryoconcentration technology. J. Food
Eng. 85, 6572.
Aider, M., de Halleux, D., Melnikova, I., 2008. Gravitational and
microwave-assisted thawing during milk whey
cryoconcentration. J. Food Eng. 88, 373380.
Aider, M., de Halleux, D., 2009. Cryoconcentration technology in
the bio-food industry: principles and applications. LWT Food
Sci. Technol. 42, 679685.
Aider, M., Ounis, W.B., 2012. Skim milk cryoconcentration is
affected by the thawing mode: gravitational vs.
microwave-assisted. Int. J. Food Sci. Technol. 47, 195202.
Akyurt, M., Zaki, G., Habeebullah, B., 2002. Freezing phenomena
in icewater systems. Energy Convers. Manage. 43, 17731789.
Boss, E.A., Filho, R.M., de Toledo, E.C.V., 2004. Freeze drying
process: real time model and optimization. Chem. Eng.
Process. 43, 14751485.

Burdo, O.G., Kovalenko, E., Kharenko, D., 2008. Intensication of

the processes of low-temperature separation of food
solutions. Appl. Therm. Eng. 28, 311316.
Esquivel, P., Jimnez, V.M., 2012. Functional properties of coffee
and coffee by-products. Food Res. Int. 46, 488495.
Gao, W., Habib, M., Smith, D.W., 2009. Removal of organic
contaminants and toxiciy from industrial efuents using
freezing processes. Desalination 245, 108119.
Habib, B., Farid, M., 2006. Heat transfer and operating conditions
for freeze concentration in a liquidsolid uidized bed heat
exchanger. Chem. Eng. Process. 45, 698710.
Huige, N.J.J., Thijssen, H.A.C., 1972. Production of large crystals by
continuous ripening in a stirrer tank. J. Cryst. Growth,
483487.
Jot, T., Laffargue, A., Descroix, F., Doulbeau, S., Bertrand, B.,
Kochko, A.D., Dussert, S., 2010. Inuence of environmental
factors, wet processing and their interactions on the
biochemical composition of green Arabica coffee beans. Food
Chem. 118, 693701.
Miyawaki, O., Kato, S., Watabe, K., 2012. Yield improvement in
progressive freeze-concentration by partial melting of ice. J.
Food Eng. 108, 377382.
Miyawaki, O., Liu, L., Shirai, Y., Sakashita, S., Kagitani, K., 2005.
Tubular ice system for scale-up of progressive
freeze-concentration. J. Food Eng. 69, 107113.
Nakagawa, K., Maebashi, S., Maeda, K., 2010a. Freeze-thawing as
a path to concentrate aqueous solution. Sep. Purif. Technol.
73, 403408.
Nakagawa, K., Nagahama, H., Maebashi, S., Maeda, K., 2010b.
Usefulness of solute elution from frozen matrix for
freeze-concentration technique. Chem. Eng. Res. Des. 88,
718724.
Nothnagl, M., Kosiba, A., Larsen, R., 2004. Predicting the effect of
irradiance and temperature on the ower diameter of
greenhouse grown Chrysanthemum. Sci. Horti. 99,
319329.
Okawa, S., Ito, T., Saito, A., 2009. Effect of crystal orientation on
freeze concentration of solutions. Int. J. Refrig. 32,
246252.
Pardo, J.M., Suess, F., Niranjan, K., 2002. An investigation into the
relationship between freezing rate and mean ice crystal size
of coffee extracts. Trans. IChemE 80, 176182.
Rattanadecho, P., 2004. Theoretical and experimental
investigation of microwave thawing of frozen layer using a
microwave oven (effects of layered congurations and layer
thickness). Int. J. Heat Mass Transfer 47, 937945.
Ravents, M., Hernndez, E., Auleda, J., Ibarz, A., 2007.
Concentration of aqueous sugar solutions in a multi-plate
cryoconcentrator. J. Food Eng. 79, 577585.
Ryyniinen, S., 1995. The electromagnetic properties of food
materials: a review of the basic principles. J. Food Eng. 26,
409429.
Snchez, J., Ruiz, Y., Auleda, J.M., Hernandez, E., Ravents, M.,
2009. Review: freeze concentration in the fruit juices industry.
Food Sci. Technol. Int. 15, 303315.
Snchez, J., Hernndez, E., Auleda, J.M., Ravents, M., 2011a.
Review: freeze concentration technology applied to dairy
products. Food Sci. Technol. Int. 17, 513.
Snchez, J., Hernndez, E., Auleda, J.M., Ravents, M., 2011b.
Freeze concentration of whey in a falling-lm based pilot
plant: process and characterization. J. Food Eng. 103,
147155.
Wang, R., Zhang, M., Mujumdar, A.S., Jiang, H., 2011. Effect of salt
and sucrose content on dielectric properties and microwave
freeze drying behavior of re-structured potato slices. J. Food
Eng. 106, 290297.
Wu, H., Tao, Z., Chen, G., Deng, H., Xu, G., Ding, S., 2004.
Conjugate heat and mass transfer process within porous
media with dielectric cores in microwave freeze drying.
Chem. Eng. Sci. 59, 29212928.